A set-membership approach to improved PNLMS algorithm

To improve convergence speed across different sparseness levels, an approach adapting dynamically to the level of sparseness using a new proportionate-type NLMS algorithm is researched. It can keep the fast initial convergence speed during the whole adaptation process until the adaptive filter reaches its steady state, but it may come at the expense of a slight increase in the computational complexity per update. For this reason, the idea of improving proportionate adaptation combined with the framework of set-membership filtering in an attempt to reduce computational complexity of algorithm is presented. Because of fewer coefficients updated in the new algorithm, the computational effort decreases significantly. Simulation results show the proposed algorithm obtains a faster convergence rates than IPNLMS algorithm in sparse circumstances, and has the same performance with their conventional counterparts for situation of dispersive channel.